Collagen is a protein found in many tissues in animals and humans, ultimately making up about 30% of the entire body’s protein content. Di Lullo 2002 While there are over 20 different types of collagen, the main types found in humans are I, II, and III. Type I is the most common and is found in tissues like bones, ligaments, tendons, and skin. Conversely, type II collagen is predominantly found in joint cartilage, while type III collagen is found in skin, blood vessels, and organs. Exposito 2010
- Collagen is a protein found in many foods like animal muscle, connective tissue, egg whites, and spirulina. It is also available in supplemental form. It is not a particularly good protein source for building muscle because it lacks the essential amino acid tryptophan, and its digestibility is much lower than other protein sources such as meat, dairy products, eggs, whey, etc.
- There are numerous types of collagen in the body; Type I is the most common and is found in tissues like bone, ligaments, tendons, and skin. Tendons are comprised of approximately 65-80% collagen protein.
- Tendinopathy is an umbrella term describing tendon-related pain and/or dysfunction. While the exact cause is unknown, tendinopathies are thought to result from impaired tendon healing from acute insults, repetitive stress, or excessive workload. Overall, there is some low-quality evidence suggesting supplemental collagen may improve tendinopathy and/or joint pain in the short term, but not in the long term.
Changes in collagen have been associated with altered function of various tissues. For example, the loss/degeneration of collagen in the skin tends to reduce elasticity and promote the development of wrinkles. Given the ubiquity of collagen in the human body, it should come as no surprise that many claim supplemental collagen can improve the function of these collagen-containing tissues, e.g. reduce wrinkles, improve joint pain, etc. In this month’s research review, we’ll get the scoop on collagen protein!
Where Does Collagen Protein Come From?
Supplemental collagen protein is sourced from ground up parts from fish, chicken, pig, cow, etc. While collagen protein contains up to 19 different amino acids, with relatively high concentrations of hydroxyproline, glycine, and proline, it is a relatively poor dietary protein source for two reasons:
- Collagen protein contains zero tryptophan, which is one of the 9 Essential Amino Acids (EAAs). High quality proteins have a high EAA content are are also easily digestible.
- It is a low quality protein when compared to whey, soy, and other proteins.
One way to assess the quality of a protein is to calculate the Protein Digestibility Corrected Amino Acid Score (PDCAAS), which compares the amount of the “limiting” amino acid in the test protein the amount of the same amino acid in 1 gram of the reference protein (egg or milk). The values is then multiplied by the test protein’s digestibility. Gropper 2012 In the case of collagen protein, the “limiting” amino acid is tryptophan, which is not present at all. Thus, the PDCAAS of collagen is zero. If a collagen protein contained added tryptophan however, the PDCAAS score would be 0.39. For comparison, the PDCAAS of whey protein is 1.00, the maximum score possible. Phillips 2016
It does appear that collagen’s higher concentrations of hydroxyproline, glycine, and proline do result in higher blood levels of those amino acids after oral ingestion. Skov 2019 However, these amino acids do not choose where they are used in the body, much the same way other sources of dietary protein can’t select where they are used. Like other protein sources (e.g. whey, soy, etc.), collagen can’t be absorbed by the body in its whole form. Rather, it must be broken down into its constituent peptides and amino acids prior to absorption. In other words, all dietary protein can be used for producing other protein in the body such as muscle, collagen, etc. once they are broken down into shorter strings of amino acids (e.g. peptides) or single amino acids so they can be absorbed in the small intestine. There are collagen peptides or hydrolyzed collagen supplements sold commercially and these supplements are absorbed intact. Still, they don’t get to choose where they’re used once absorbed. Finally, there are many foods that are naturally rich in collagen such as bone broth, animal muscle or connective tissue, egg whites, and spirulina. Overall, collagen is just another source of amino acids, albeit a relatively poor one compared to other dietary protein sources.
Connecting the Dots Between Tendinopathy and Collagen Protein
As mentioned above, tendons have a high collagen content – about 65-80% collagen (mostly type I). Kannus 2000 Tendinopathy, which is an umbrella term describing tendon pain and/or dysfunction, is thought to result from impaired tendon healing after an acute overload, repetitive stress, or other insult. Charnoff 2019 One of the lines of thinking behind supplementing collagen is that it may improve the impaired tendon healing response associated with tendinopathy, which itself may be due to impaired collagen synthesis or repair.
To explore this relationship, we’ll take a look at a recent study investigating the effect of collagen supplementation combined with an exercise program on mid-portion Achilles tendinopathy. The Achilles tendon connects the muscles of the lower limb, e.g. the plantaris, gastrocnemius, and soleus, to the calcaneus (e.g. the heel bone). Patients with mid-portion Achilles tendinopathy typically present with pain 2-6 cm above the calcaneus (e.g. the “heel bone”), whereas insertional Achilles tendinopathy has pain at the site of insertion of the tendon on the calcaneus. This is a common issue in both elite and recreational athletes, who have a 24% and 9% lifetime risk of developing Achilles tendinopathy. Olewnik 2018 Li 2016 Achilles tendinopathy also tends to respond relatively well to conservative (i.e. non-surgical) management, but the role of collagen supplementation is currently under debate.
This paper compared the effects of collagen supplementation and exercise with placebo on 1) pain symptoms and 2) tendon vascularization.
Pain symptoms were assessed by the Victorian Institute of Sport Assessment-Achilles (VISA-A), which consists of eight questions about the person’s pain, function, and activity tolerance with respect to their Achilles tendon. The VISA-A is scored 0-100, where 100 represents a person who is completely asymptomatic and a lower score suggests increased pain and limitations in function; 70 is said to represent the maximum score for a person with Achilles tendinopathy.
The VISA-A is also sensitive for detecting changes in symptoms over time. It should be noted that the Minimum Clinically Important Difference (MCID) on this scale is 6.5. McCormack 2015 This means that a change of 6.5 or greater is thought to represent a clinically significant change in symptoms for a patient. Finally, the VISA-A also seems to give consistent scores when test-retesting the same individual or using a different tester (r = 0.93 and 0.90, respectively). Robinson 2001
With respect to tendon vascularization, researchers used Contrast-Enhanced Ultrasound (CEUS) to evaluate for new blood vessel formation in the Achilles tendon. While “normal” tendons have a relatively poor vascular supply and little to no tissue turnover during adulthood, tissue turnover is increased in those with tendinopathy. Heinemeier 2018
It may seem logical that increased tissue turnover and blood flow should happen if there’s damage to a tendon; however, some of the existing data suggest that these processes may actually be a causative factor in symptoms and tissue dysfunction. In fact, one of the intrinsic risk factors for Achilles tendinopathy or rupture of the Achilles tendon in individuals with Achilles pain is increased vascularity. Palazon-Bru 2019 Kannus 1991
Mechanistically, it appears much of the increased tissue turnover in tendinopathy seems to involve new blood vessels or tendon materials, however both tend to have disorganized structures. Specifically, the blood vessels tend to be somewhat “leaky” and not actually carry much blood to the relevant structures of the tendon. This ultimately leads to an anaerobic environment, as there is a lot of energy being used for cellular turnover, but not much oxygen-carrying blood being delivered to the tissues.
One growth factor we’d expect to see cells that are operating in a hypoxic or low oxygen environment produce is hypoxia-inducible factor-1α (HIF-1α). Indeed, when you look at tissue samples of tendons in those with tendinopathy or rupture tendons, there is an elevated expression of HIF-1α and its target genes. Pufe 2005 Another study that compared the amount of new blood vessel formation and pain in the Achilles tendon are found that 97.3% of those with symptomatic Achilles tendons had a substantial amount of new blood vessel formation, which was also correlated to the individuals’ pain scores. Yang 2012
In summary, there is uncertainty regarding the role of the vasculature in the processes underlying tendon degeneration, tendon rupture, and the symptoms being experienced. This study is using ultrasound to look at vascularization changes in the Achilles tendon over time. At this time, it appears the increased turnover and blood vessel formation may be associated with an increase in symptoms, though the clinical relevance of this finding is unclear. In other words, at this time we don’t have a strong reason to significantly change the management of tendinopathy based on tendon vascularization alone.
A total of 20 patients with Achilles tendinopathy were included in this study. Seven of the 20 individuals had Achilles tendinopathy on one side (i.e. unilateral), whereas 13 subjects had Achilles tendinopathy on both sides (i.e. bilateral). There were significant differences in symptom duration amongst participants, with some individuals having tendinopathy symptoms for less than 1 month and others having symptoms for close to 30 years. Unfortunately, the data were not reported in a way where we can tell whether the duration of symptoms affected outcomes. Baseline characteristics can be seen in Figure 1.
VISA-A questionnaires and ultrasound measurements were collected for both right and left Achilles tendon separately at the start of the study and at both 3 and 6 months.
Additionally, blood samples to determine total cholesterol, triglycerides, and uric acid were obtained at the same time points to rule out metabolic conditions like familial hypercholesterolemia and gout as contributing to tendinopathy. Based on the lab results, no individuals had evidence of these conditions.
With respect to the intervention, the study used a double-blinded placebo-controlled crossover design. In simple terms, neither the subjects nor the researchers knew who was getting the placebo and who was getting the collagen supplement (e.g. it was “double” blinded and placebo-controlled). After 3 months, the subjects who received the collagen supplement initially started receiving the placebo, whereas those who were receiving the placebo initially now started getting the collagen supplement (the “crossover” phase). The subjects were asked at 3 and 6 months whether they believed they were getting the placebo or the collagen supplement in order to assess for inadvertent “un-blinding” that could confound results.
The collagen supplement itself was 2.5 grams of hydrolyzed collagen peptides taken twice daily, with one dose taken 30 minutes prior to exercise. Conversely, the placebo contained 2.5 grams of maltodextrin and had the same packaging, taste, and texture as the collagen supplement. Adherence to intake of the collagen supplement and/or placebo was 89 ± 15% and 91 ± 9%, respectively, over the 6-month intervention period.
Finally, the exercise program combined elements of eccentric calf-strengthening and running over the 6 month study period. The study reports the following instructions:
“The eccentric exercise started with a calf raise using the unaffected limb (or both limbs if the symptoms were bilateral), followed by an eccentric drop using the injured leg. The exercise was performed with both the knee straight and [with] the knee bent to target the gastrocnemius and soleus muscles respectively. Over a 6-month period, participants were instructed to perform 2 × 90 repetitions daily, despite the presence of pain. All participants were instructed to avoid weight-bearing sporting activities for the first four weeks. When participants reported less than 2 out of 10 pain on single leg hopping, they were allowed to start with low-intensity running exercises. “
Based on the available exercise diaries, both groups showed similar adherence at 84% and 78%, respectively, to the twice-daily calf-strengthening exercise program over the 6 month study period.
VISA-A / Pain
At 3 months, the group receiving the collagen supplement improved their VISA-A scores from an average of 58.5 to 60.8, whereas the group receiving the placebo improved from an average of 55.9 to 62.8. As previously noted, the minimum clinically important difference (MCID) for the VISA-A score is 6.5 points, which means that only the placebo group saw a clinically meaningful improvement when comparing the average starting and ending VISA-A scores.
When the authors reanalyzed the data using a method called “linear mixed modeling”, they reported a 12.6 point improvement in the collagen-receiving group and a 5.3 point improvement in the placebo-receiving group during the first 3 months of the study.
During the second three-month period, the linear mixed modeling analysis found a 5.9 point improvement in the group who previously received the collagen supplement, but who then crossed over to placebo. During this same time period, the group previously getting a placebo, but who was now getting the collagen supplement showed a 17.7 point improvement based on linear mixed modeling.
Overall, both groups saw similar improvements, 18.5 and 23.0 points respectively, in their VISA-A scores between baseline and six months. Interestingly, the group receiving the placebo first showed a greater improvement in VISA-A scores, which was statistically significant. Finally, 60% (6/10) and 55% (5/9) of the individuals correctly guessed that they were getting the collagen supplement, whereas only 40% (4/10) and 44% (4/9) correctly guessed they were getting the placebo.
Both groups showed a significant decrease in tendon vascularization during the study period, however there was no significant difference between those receiving the placebo or the collagen supplement at any time. This suggests the improvement in vascularization was unrelated to supplementation and may have been due to the exercise protocol or simply due to natural history (i.e., it may have occurred over time regardless of what was done in the study).
There were no changes in average serum levels of total cholesterol, triglycerides and uric acid during the study period.
What’s the Take Home Message?
This article provides conflicting evidence on the effect of collagen supplementation on mid-portion Achilles tendinopathy. As mentioned above, when comparing the average VISA-A scores from baseline to 3 months, the placebo group seemed to outperform those receiving the collagen supplement. Only through additional statistical techniques (linear mixed modeling, a technique used to normalize heterogenous data), was the difference in favor of collagen supplementation. It’s not clear what this actually means, however, as no raw individual or group data were provided. When coupled with the very small sample size (n=20), this makes it difficult to determine whether the linear-mixed-modeling-reported improvements are meaningful and can be generalized for application to the general public.
One finding that suggests the collagen supplement may not be uniquely beneficial is that those who received the placebo for the first three months and the collagen supplement for the second three months did better than those who received the supplement in the reverse order. If the supplement was uniquely beneficial we’d expect those with worse VISA-A scores to see a more robust improvement than those with better VISA-A scores. However, the VISA-A scores were, on average, universally worse at the beginning of the study when compared to the time point when the crossover occurred. The data reported – even when using linear mixed modeling – suggest that there was a smaller improvement with the collagen supplement when the scores were worse than with placebo.
Additionally, 60% (6/10) and 55% (5/9) of the individuals correctly guessed that they were getting the collagen supplement, whereas only 40% (4/10) and 44% (4/9) correctly guessed they were getting the placebo. This suggests that if the linear mixed modeling data correctly identified a greater improvement in symptoms, that there may have been a placebo-induced “positive expectation” effect taking place. In other words, when individuals suspected they were getting the “active” supplement, they had better outcomes than when they thought they were getting the placebo. Focusing on surrogate outcomes like various biomarkers rather than clinically relevant (and validated) outcomes is also problematic.
It is unfortunate that this study used a maltodextrin placebo rather than another protein source. Recall from the introduction that collagen is a protein that, like all others, consists of a chain of amino acids that cannot be absorbed in its whole form. This is not a unique problem, however, as much of the existing research on the benefits of collagen supplementation with a control group uses either water or sugar-based placebo (e.g. maltodextrin). Moreover, nearly all of the research – including the majority of papers that do not have an actual control group – has focused on collagen supplements and not food sources that are rich in collagen or food products containing collagen as a functional additive.
It may be that the ingestion of amino acids from any dietary protein – not that the specific amino acids derived from a collagen supplement – may be related to observed improvements. Remember that peptides and amino acids do not get to choose the tissues where they are ultimately used, e.g. the muscle, bone, tendons, skin, organs, etc. Rather, the body will direct these building blocks towards where they’re needed, regardless of source.
My impression of the current study is that the benefit that could be reasonably attributed to collagen supplementation is either very small or non-existent. Similar results are shown in other studies addressing the efficacy of collagen supplementation on joint function. Lugo 2013 However, currently there’s little evidence on whether collagen supplements can decrease joint injuries or accelerate athletes’ return to play after injury. Heaton 2017 Rawson 2018 Finally, a 2018 meta analysis reviewed 69 eligible studies (including collagen) on nutrition supplements and their effect on joint pain. The authors concluded that,
“No supplements were identified with clinically important effects on pain reduction at long term.” Liu 2018
Overall, I don’t think the current evidence suggests that there is a reliable effect of supplemental collagen protein on tendinopathy. More well-controlled studies investigating the unique effects of collagen supplements (ideally versus other protein sources) on tendinopathy are needed.
Di Lullo GA, Sweeney SM, Korkko J, Ala-Kokko L, San Antonio JD. Mapping the ligand-binding sites and disease-associated mutations on the most abundant protein in the human, type I collagen. J Biol Chem. 2002;277(6):4223-4231.
Gropper, Sareen S.. Advanced Nutrition and Human Metabolism (p. 239). Cengage Textbook. Kindle Edition.
Phillips SM. The impact of protein quality on the promotion of resistance exercise-induced changes in muscle mass. Nutr Metab (Lond). 2016;13:64. Published 2016 Sep 29. doi:10.1186/s12986-016-0124-8
Skov K, Oxfeldt M, Thøgersen R, Hansen M, Bertram HC. Enzymatic Hydrolysis of a Collagen Hydrolysate Enhances Postprandial Absorption Rate-A Randomized Controlled Trial. Nutrients. 2019;11(5):1064. Published 2019 May 13. doi:10.3390/nu11051064
Kannus, P. (2000), Structure of the tendon connective tissue. Scandinavian Journal of Medicine & Science in Sports, 10: 312-320. doi:10.1034/j.1600-0838.2000.010006312.x
Charnoff J, Naqvi U. Tendinosis (Tendinitis) [Updated 2019 Feb 4]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2019 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK448174/
Olewnik Ł, Wysiadecki G, Podgórski M, Polguj M, Topol M. The Plantaris Muscle Tendon and Its Relationship with the Achilles Tendinopathy. Biomed Res Int. 2018
Li HY, Hua YH. Achilles Tendinopathy: Current Concepts about the Basic Science and Clinical Treatments. Biomed Res Int. 2016
J M Robinson, J L Cook, C Purdam, P J Visentini, J Ross, N Maffulli, J E Taunton, K M Khan, for the Victorian Institute of Sport Tendon Study Group. The VISA-A questionnaire: a valid and reliable index of the clinical severity of Achilles tendinopathy. Br J Sports Med 2001;35:335-341.
McCormack J, Underwood F, Slaven E, Cappaert T. THE MINIMUM CLINICALLY IMPORTANT DIFFERENCE ON THE VISA-A AND LEFS FOR PATIENTS WITH INSERTIONAL ACHILLES TENDINOPATHY. Int J Sports Phys Ther. 2015;10(5):639–644.
Heinemeier KM, Schjerling P, Øhlenschlæger TF, et al Carbon-14 bomb pulse dating shows that tendinopathy is preceded by years of abnormally high collagen turnover. Faseb J 2018;32:4763–75.
Pufe T, Petersen WJ, Mentlein R, et al. The role of vasculature and angiogenesis for the pathogenesis of degenerative tendons disease. Scand J Med Sci Sports 2005;15:211–22.
Palazón-Bru A, Tomás-Rodríguez MI, Mares-García E, Gil-Guillén VF. A Reliability Generalization Meta-Analysis of the Victorian Institute of Sport Assessment Scale for Achilles Tendinopathy (VISA-A). Foot Ankle Int. 2019 Apr;40(4):430-438.
Kannus P, Jozsa L: Histopathological changes preceding spontaneous rupture of a tendon. A controlled study of 891 patients. J Bone Joint Surg Am. 1991, 73: 1507-1525.
Yang X, Coleman DP, Pugh ND, Nokes LD. The volume of the neovascularity and its clinical implications in achilles tendinopathy. Ultrasound Med Biol. 2012 Nov;38(11):1887-95.
Lugo JP, Saiyed ZM, Lau FC, et al. Undenatured type II collagen (UC-II®) for joint support: a randomized, double-blind, placebo-controlled study in healthy volunteers. J Int Soc Sports Nutr. 2013;10(1):48.
Heaton LE, Davis JK, Rawson ES, et al. Selected in-season nutritional strategies to enhance recovery for team sport athletes: a practical overview. Sports Med. 2017;47(11):2201-2218.
Rawson ES, Miles MP, Larson-Meyer DE. Dietary supplements for health, adaptation, and recovery in athletes. Int J Sport Nutr Exerc Metab. 2018;28(2):188-199.
Liu X, Machado GC, Eyles JP, Ravi V, Hunter DJ. Dietary supplements for treating osteoarthritis: a systematic review and meta-analysis. Br J Sports Med. 2018;52(3):167-175.